Testing completion depends on risk assessment and acceptable quality thresholds for the system being tested, not on running all planned tests (A), time limits (B), or fixing all faults (C) which may be impractical. Option E correctly states testing stops when risks are mitigated to acceptable levels.